Thanks to Harry for the excellent summary and comments. His two comments really amount to a single challenge: Since a smaller-than-universal sample of the population could deliberate or vote with as much epistemic value as the universal franchise, doesn’t epistemic proceduralism end up endorsing non-democratic political institutions after all?
It’s one challenge, but with two parts. So, first, is it true that a sample could be trusted (beyond qualified disagreement) to deliberate with as good effect as if the whole adult population were enfranchised? Then, second, if so, would deliberation and/or voting by a sample count as democratic?
There’s a third question in play, raised when Harry writes, “Democracy is, in part, justified because we have an obligation to allow other people to play a full role in determining how our collective affairs will go, precisely because they have a stake in how things will go.” Here the question is not what counts as democratic, but whether we are obligated to enfranchise all adult citizens.
Starting with that third question, we should notice that it isn’t quite the same as any of the questions I address, which is not to say that it isn’t just as important. But it isn’t either the question of legitimacy or authority as I understand them. Apart from whether only a democratic regime can permissibly enforce its laws, or place moral obligations on citizens to obey them, there might in any case be an obligation on all of us to produce and maintain democratic institutions. Technically, I haven’t taken any stand either way on that question. But Harry might mean to suggest also that legitimacy and/or authority depend on democracy, in which all adult citizens can vote, “just because they have a stake in how things will go.”
I admit that this is intuitively tempting. It conjures an analogy with individual liberty: I get to decide what I do (in a certain range of cases) simply because I have a special interest in my own actions, and not because I will make good decisions. So we can imagine a similar principle of liberty that protects the self-regarding actions of “the people.” Harry doesn’t say this; I’m just trying to bring to the fore a way of seeing the matter than tempts us toward his claim that voters should get to decide because they have a stake. This tempting analogy fails, I think. The usual liberal reasons for letting an individual control a certain sphere of choice treat the case for such liberty as being much weaker when the individual’s actions affect others in certain ways. But giving me the power to vote gives me the power to affect (with a small probability individually, but with determinacy in large numbers) the lives of people who would choose differently. Any democrat must think this is justified, but it can’t be by analogy with the liberty of individual action, since that liberty finds profound limits when the choices limit the liberty of others. I don’t know of any strong argument that grounds the authority, legitimacy, or moral necessity of democracy in a principle about those with a stake having a right to decide. In fact, that isn’t a principle that we generally accept in other contexts. Students have a stake, but they don’t get to decide the grading curve. Soldiers have a stake but they don’t get to decide on the battle assignments. I do think that citizens should decide (at least indirectly) on laws and policies, but not on the simple basis of their having a stake in the decisions.
Even so, part of Harry’s point there is different. I take him to be asserting (quite apart from the stakeholder idea) that an arrangement in which not all adult citizens are enfranchised does not count as democratic. He adds the moral principle about stakeholding, I take it, because he’s interested in a morally important concept of democracy, not in a sterile debate about the definition of a word or the extension of a concept. Me too. I define democracy as a political arrangement in which laws and policies are, directly or indirectly, actually collectively authorized by the people subject to them. This clearly doesn’t mean authorized by each and every subject. No theory of democracy could deliver that, since a unanimity rule is universally rejected for large political organizations like nation-states. A voter who loses does not, herself, in that case, authorize the resulting laws and policies. So my definition requires that, if democracy is even remotely possible, the people subject to the laws can count has having collectively authorized them even when not every individual has. I don’t offer a theory of collective authorization, except to speak to certain important cases: I assume that majority rule is not automatically a failure of collective authorizaion just because there is, in each decision, a minority who is outvoted. And I assume that a random choice of policies is not a case of anyone authorizing anything (unless they authorized that procedure, a complication I can’t go into here).
If an outvoted minority is not inconsistent with collective authorization, even though they positively voted against the decision that passed, then we can keep an open mind about whether the fact that some are not chosen to vote at all in a random lottery disqualifies it as a case of collective authorization. You might say, at least in the normal case they all voted. But in fact we know that in any given election a large fraction of eligible voters in modern democracies do not vote. Maybe if too many stay home we would not count it as collective authorization by all. But we don’t require that all vote for it to be collective authorization. Whatever fraction of voters is enough, consider a lottery that chooses that many voters instead. The same number vote in both cases. So the general voting procedure does not have that advantage over the lottery.
What I take this to show is that it isn’t obvious, and there is no clear reason on the table for believing, that people can only collectively authorize a decision if they all have a vote. Since the actual collective authorization of laws by the people subject to them is what I have in mind by democracy, democracy does not necessarily rule out choosing the voters by lottery. As to whether democracy, so conceived, can ground legitimacy and authority, that’s the argument of the book as a whole.
We can certainly ask whether some sample of all voters might, so far as public reason is concerned, be epistemically just as good. If there were no qualified objection to that proposition, then my account says that the smaller deliberating or voting group would contribute just as much to political legitimacy and authority as a larger deliberation would.
One point that counts against this possibility is that there are potentially at least as many relevant reasons and considerations as there are individuals. That is some pressure toward having everyone deliberate. Here’s another consideration: as I know from the case of my 18 year old daughter, people who know they will actually vote are often more attentive to politics and more motivated to find out about it, in order to prepare for their vote. I suspect the same might be true about knowing you will be among those in the deliberating sample. If each citizen was only in the sample group occasionally she would lose much of this motivation and attentiveness. I am not appealing to the education of citizens as an independent value. My point is that citizens who lose this attentiveness will be less competent voters and deliberators when their turn comes. That would be a reason for giving everyone a vote each time, even though at any given time we could do just as well by taking a sample. The point is to build up the competence of future samples, but since there are always more future samples to educate, there is never actually any time at which only a sample should be consulted.
If sampling counts as collective authorization (which I’m not sure about), then I count it as democratic. (And if it is at least as good as having everyone vote, then my theory recommends it.) If someone doesn’t count it as democratic because they define democracy so as to preclude sampling, then, if samples would perform as well, my argument does not necessarily recommend democracy in that particular sense. The argument is that the epistemically best arrangement so far as could be determined beyond qualified disagreement would be an arrangement that would also count as actual collective authorization of laws by the people. It’s in that particular meaning of democracy that it is an argument for democracy. As an aside, I’m inclined to think that even if a sample would do just as well, and even if this was beyond qualified objection, there might be actual objection that is worth avoiding by avoiding sampling. If there’s a gain in social peace, and no loss, why not have everyone vote? But if sampling were actually much better, then such resistance would be less dispositive.
So, as the sampling question shows, epistemic proceduralism can fairly be pressed on whether it fixes the set of deliberators and voters in a plausible place. All adults? Why not fewer (e.g., sampling)? So we can also ask, with Ben, why not more (e.g., foreigners)? Again, I don’t regard it as a desideratum that the theory should identify as eligible voters all and only adult citizens. But it would be intuitively implausible if citizens had no more claim to the franchise than foreigners. I haven’t thought about this question much, but here’s the background assumption I’ve been operating with: roughly speaking, the set of citizens will be a set of people whose stake in the collective decision, and experience and familiarity with past decisions in that community, gives them a special epistemic advantage. This effect could be especially strong if, for example, there is a strong distinctive culture that covers that same group, but it would have some relevance even if there isn’t. Of course, this would not draw a bright line. I would not think it fatal if some small number of foreign citizens could, within public reason, be held to have so much to contribute that they should be somehow invited to contribute. (Many will rebel at this idea, but many objections would depend on denying the epistemic premise about foreign voters. Suppose in including them would lead to better decisions.) We need to keep in mind, however, that there is likely to be a lot of qualified disagreement about what the criteria for such foreign voters would be, just as there is for any epistocratic proposal. This is about as much as I’m prepared to say about the question, except to observe that other theories will face the same issue. For example, a fair-proceduralist view might be pressed to explain why non-citizens who have a stake in the decisions shouldn’t also have a fair role. A utilitarian view of democracy would have to explain why we should think utility is best served by letting only citizens of a democracy vote for it’s laws and leaders. Many will assert a simple right of a person to have a role in the making of laws that apply to him. But if we ask why we should think there is a right of that precise shape, similar difficulties arise.
Ben: Benevolent dictatorship might perform better than democracy, but it depends on the dictator, and I claim there would always be qualified disagreement about that. As for other non-democratic forms of government and the chance that they would perform better than democracy: I think that if the case could be made beyond qualified disagreement, this would defeat democracy’s claims to authority and legitimacy. Democracy, on my view, is not a first principle, so there are imaginable conditions in which it would be inappropriate. Now, which non-democratic form of government would perform better with respect to avoiding the primary bads in, say the U.S. or the U.K. at this historical moment? Or in which other countries is this the case? If it is the case for some, then I suspect they should not be democratic, but we’d have to look at the case and see how big a bullet that would be to bite.
Jonathan: That interesting proposal of a lottery-chosen specially-trained panel of voters looks a lot like epistocracy in the relevant sense. But I see how the lottery aspect helps. Part of the reasonable worry about epistocracy of the educated stems from the factors that select who gets the education. That factor is removed in your proposal. So the question would be about qualified disagreement over what counts as good training. I suspect there’s not much that’s beyond qualified disagreement there. But if there is, my theory says it is the arrangement that would have legitimate authority. Would it count as democratic? That takes us back to the issues at the beginning of this post.